The ability to manipulate the composition of crop seeds, particularly the content and composition of seed oils, has important applications in the agricultural industries, relating both to processed food oils and to oils for animal feeding. Seeds of agricultural crops contain a variety of valuable constituents, including oil, protein and starch. Industrial processing can separate some or all of these constituents for individual sale in specific applications. For instance, nearly 60% of the U.S. soybean crop is crushed by the soy processing industry. Soy processing yields purified oil, which is sold at high value, while the remainder is sold principally for lower value livestock feed (U.S. Soybean Board, 2001 Soy Stats). Canola seed is crushed to produce oil and the co-product canola meal (Canola Council of Canada). Nearly 20% of the 1999/2000 U.S. corn crop was industrially refined, primarily for production of starch, ethanol and oil (Corn Refiners Association). Thus, it is often desirable to maximize oil content of seeds. For instance, for processed oilseeds such as soy and canola, increasing the absolute oil content of the seed will increase the value of such grains. For processed corn it may be desired to either increase or decrease oil content, depending on utilization of other major constituents. Decreasing oil may improve the quality of isolated starch by reducing undesired flavors associated with oil oxidation. Alternatively, in ethanol production, where flavor is unimportant, increasing oil content may increase overall value. In many fed grains, such as corn and wheat, it is desirable to increase seed oil content, because oil has higher energy content than other seed constituents such as carbohydrate. Oilseed processing, like most grain processing businesses, is a capital-intensive business; thus small shifts in the distribution of products from the low valued components to the high value oil component can have substantial economic impacts for grain processors.
Biotechnological manipulation of oils can provide compositional alteration and improvement of oil yield. Compositional alterations include high oleic soybean and corn oil (U.S. Pat. Nos. 6,229,033 and 6,248,939), and laurate-containing seeds (U.S. Pat. No. 5,639,790), among others. Work in compositional alteration has predominantly focused on processed oilseeds but has been readily extendable to non-oilseed crops, including corn. While there is considerable interest in increasing oil content, the only currently practiced biotechnology in this area is High-Oil Corn (HOC) technology (DuPont, U.S. Pat. No. 5,704,160). HOC employs high oil pollinators developed by classical selection breeding along with elite (male-sterile) hybrid females in a production system referred to as TopCross. The TopCross High Oil system raises harvested grain oil content in maize from about 3.5% to about 7%, improving the energy content of the grain.
While it has been fruitful, the HOC production system has inherent limitations. First, the system of having a low percentage of pollinators responsible for an entire field's seed set contains inherent risks, particularly in drought years. Second, oil contents in current HOC fields have plateaued at about 9% oil. Finally, high-oil corn is not primarily a biochemical change, but rather an anatomical mutant (increased embryo size) that has the indirect result of increasing oil content. For these reasons, an alternative high oil strategy, particularly one that derives from an altered biochemical output, would be especially valuable.
The most obvious target crops for the processed oil market are soy and rapeseed, and a large body of commercial work (e.g., U.S. Pat. No. 5,952,544; PCT application WO9411516) demonstrates that Arabidopsis is an excellent model for oil metabolism in these crops. Biochemical screens of seed oil composition have identified Arabidopsis genes for many critical biosynthetic enzymes and have led to identification of agronomically important gene orthologs. For instance, screens using chemically mutagenized populations have identified lipid mutants whose seeds display altered fatty acid composition (Lemieux et al., 1990; James and Dooner, 1990). T-DNA mutagenesis screens (Feldmann et al., 1989) that detected altered fatty acid composition identified the omega 3 desaturase (FAD3) and delta-12 desaturase (FAD2) genes (U.S. Pat. No. 5,952,544; Yadav et al., 1993; Okuley et al., 1994). A screen which focused on oil content rather than oil quality, analyzed chemically-induced mutants for wrinkled seeds or altered seed density, from which altered seed oil content was inferred (Focks and Benning, 1998). Another screen, designed to identify enzymes involved in production of very long chain fatty acids, identified a mutation in the gene encoding a diacylglycerol acyltransferase (DGAT) as being responsible for reduced triacyl glycerol accumulation in seeds (Katavic V et al, 1995). It was further shown that seed-specific over-expression of the DGAT cDNA was associated with increased seed oil content (Jako et al., 2001).
Activation tagging in plants refers to a method of generating random mutations by insertion of a heterologous nucleic acid construct comprising regulatory sequences (e.g., an enhancer) into a plant genome. The regulatory sequences can act to enhance transcription of one or more native plant genes; accordingly, activation tagging is a fruitful method for generating gain-of-function, generally dominant mutants (see, e.g., Hayashi et al., 1992; Weigel D et al. 2000). The inserted construct provides a molecular tag for rapid identification of the native plant whose mis-expression causes the mutant phenotype. Activation tagging may also cause loss-of-function phenotypes. The insertion may result in disruption of a native plant gene, in which case the phenotype is generally recessive.
Activation tagging has been used in various species, including tobacco and Arabidopsis, to identify many different kinds of mutant phenotypes and the genes associated with these phenotypes (Wilson et al., 1996, Schaffer et al., 1998, Fridborg et al., 1999; Kardailsky et al., 1999; Christensen S et al. 1998).